Floats

Intro

Analysis of floats deployed at N2S4 (n0647) and N3S6 (n0847, n0846) occupied S6 from 9/13-9/17

• Input data, which were the level 2 processed data, were pulled from the NAAMES data page
• MLDs were calculated using the N² buoyancy frequency and are defined as the depth below 5 m at which N² > |stdev(N²)|

Import and Wrangle Data

#pass all the filenames of everything in a float folder into a variable
filenames <- list.files("Input/n0847/",pattern = "*.csv", full.names = T) 
profile.names <- list.files("Input/n0847/",pattern = "*.csv", full.names = F) 
profile.number <- gsub('.csv', '', profile.names) %>% 
  gsub('n0847.', '', .) %>% 
  as.numeric(.)

#now go through all the files in the folder, pick out the filenames, and pass them into the variable 'filenames'. use the file names to interate through the files, extracting the data from each csv, and combining it into one list
master <-  lapply(filenames, function(i){
  read_csv(i)
})
names(master) <- profile.number

#convert to data frame
profiles <- bind_rows(master, .id = "profile")

#save csv
# write_csv(profiles, "Input/n0847_profiles.csv")

n0647 <- read_csv("Input/n0647_profiles.csv") %>%  
  mutate(float = "n0647", Cruise = "AT34", Station = 4) %>% select(float:Station, profile, datetime:s, par, o2_c, chla_adj, fdom, cphyto, poc) %>%
  rename(o2 = o2_c, chla = chla_adj) %>% 
  filter(datetime < ymd_hms("2016-06-03 06:35:00"))

n0846 <- read_csv("Input/n0846_profiles.csv") %>%  
  mutate(float = "n0846", Cruise = "AT38", Station = 6) %>% select(float:Station, profile, datetime:s, par, o2_c, chla_adj, fdom, cphyto, poc) %>%
  rename(o2 = o2_c, chla = chla_adj) %>% 
  filter(datetime < ymd_hms("2017-09-21 00:00:00"))

n0847 <- read_csv("Input/n0847_profiles.csv") %>%  
  mutate(float = "n0847", Cruise = "AT38", Station = 6) %>% select(float:Station, profile, datetime:s, par, o2_c, chla_adj, fdom, cphyto, poc) %>%
  rename(o2 = o2_c, chla = chla_adj) %>% 
  filter(datetime < ymd_hms("2017-09-21 00:00:00"))

Calculations

MLDs and EZD

Euphotic zone depth calculated as depth at which PAR is 1% surface PAR.

#combine all the float data
naames_floats <- rbind(n0647, n0846, n0847) 

floatsNmlds <-  naames_floats %>% 
  group_by(float, profile, datetime) %>% 
  mutate(sigT = oce::swSigmaTheta(salinity = s, temperature = t, pressure = p, referencePressure = 0),
         potT = swTheta(salinity = s, temperature = t, pressure = p),
         N2 = oce::swN2(pressure = p, sigmaTheta = sigT)) %>% 
  filter(p > 5 & N2 > abs(sd(N2))) %>% 
  filter(p == min(p)) %>% #report the shallowest depth at which the above condition is met
  rename(MLD = p) %>% 
  ungroup() %>% 
  select(float:MLD) %>% 
  distinct() %>% 
  left_join(naames_floats, .) 
  
  # filter(datetime == ymd_hms("2016-05-24 12:28:04")) %>%
  
floatsNmldsNezds <- floatsNmlds %>% 
  group_by(float, profile, datetime) %>% 
  mutate(onePercSurfPAR = 0.01 * first(par)) %>% 
  group_modify(~ add_row(., par = max(.$onePercSurfPAR))) %>% 
  fill(c(Cruise, Station, lat, lon, MLD, onePercSurfPAR)) %>% 
  arrange(Cruise, Station, lat, lon, MLD, par) %>% 
  mutate(p = zoo::na.approx(p, par, na.rm = F),
         ezd = ifelse(par == onePercSurfPAR, p, NA)) %>% 
  fill(ezd, .direction = "downup") %>% 
  select(-onePercSurfPAR) %>% 
  drop_na(s) %>% 
  ungroup() %>% 
  select(float, profile, datetime, ezd) %>% 
  distinct() %>% 
  left_join(floatsNmlds, .)  %>% 
  mutate( potT = swTheta(salinity = s, temperature = t, pressure = p),
          date = ymd(as_date(datetime)), 
          time = as_hms(datetime),
          plot_date = paste(month(date, label = T), day(date), format(parse_date_time(time, c('HMS', 'HM')), '%H:%M'))) %>% 
  select(float:datetime, date, time, plot_date, Cruise:t, potT, everything())

summary(floatsNmldsNezds$ezd)  

##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##   11.19   39.08   46.87   49.71   64.11   76.98    6562

T-S plots

We’ll make T-S plots from the floats and overlay the data from the ship collected data.

ship_ctd <- read_rds("Input/ctd_data.rds") %>% 
  filter(Cruise == "AT34" & Station == 4 | Cruise == "AT38" & Station == 6) %>% 
  select(CampCN:Station, `Longitude [degrees_east]`, lat, mld, pres_db, ave_temp_c, ave_sal_psu) %>% 
  rename(datetime = Date, lon = `Longitude [degrees_east]`, p = pres_db, s = ave_sal_psu, t = ave_temp_c, MLD = mld) %>% 
  mutate( potT = swTheta(salinity = s, temperature = t, pressure = p),
          date = ymd(as_date(datetime)), 
          time = as_hms(datetime),
          plot_date = paste(month(date, label = T), day(date), format(parse_date_time(time, c('HMS', 'HM')), '%H:%M'))) %>% 
  select(CampCN:datetime, date, time, plot_date, Cruise:t, potT, everything())

s4_subset <- floatsNmldsNezds %>% filter(Cruise == "AT34", p <= 250,  datetime < ymd_hms("2016-05-31 00:00:00")) %>% mutate(type = "Float") %>% select(-c(o2:poc)) %>% distinct() %>% 
  bind_rows(., ship_ctd %>% filter(Cruise == "AT34", p <= 250) %>% mutate(type = "Ship"))

s4_subset$plot_date <- reorder(s4_subset$plot_date, s4_subset$datetime)

s4.ts <- PlotSvalbard::ts_plot(s4_subset, temp_col = "potT", sal_col = "s", color = "type", symbol_size = 3, symbol_shape = 16, symbol_alpha = 0.5, color_isopyc = "black", WM = NULL) +
  scale_color_viridis_d(begin = 0.1, end = 0.8) +
  theme_classic2(18) +
  scale_x_continuous(name = "Practical Salinity", sec.axis = sec_axis(~.,name = expression(paste("Potential Density, kg m"^-3)))) +
  scale_y_continuous(name = "Potential Temperature, ˚C") +
   theme(panel.spacing.x = unit(1, "cm"),
        axis.text.x = element_text(angle = 0),
        legend.title = element_blank(),
        legend.key.size = unit(0.4, "cm"),
        legend.position = c(0.8, 0.25),
        legend.text = element_text(size = 16),
        legend.background = element_rect(size = 0.2, linetype = "solid", color = "black"),
        legend.spacing.y = unit(0, "pt"))

s4_subset_ship <- s4_subset %>% filter(type == "Ship")

s4.ts2 <- PlotSvalbard::ts_plot(s4_subset_ship, temp_col = "potT", sal_col = "s", color = "plot_date", symbol_size = 3, symbol_shape = 16, symbol_alpha = 0.5, color_isopyc = "black", WM = NULL) +
  scale_color_viridis_d(begin = 0.1, end = 0.8) +
  theme_classic2(18) +
  scale_x_continuous(name = "Practical Salinity", sec.axis = sec_axis(~.,name = expression(paste("Potential Density, kg m"^-3)))) +
  scale_y_continuous(name = "Potential Temperature, ˚C") +
   theme(panel.spacing.x = unit(1, "cm"),
        axis.text.x = element_text(angle = 0),
        legend.title = element_blank(),
        legend.key.size = unit(0.4, "cm"),
        legend.position = c(0.8, 0.25),
        legend.text = element_text(size = 12),
        legend.background = element_rect(size = 0.2, linetype = "solid", color = "black"),
        legend.spacing.y = unit(0, "pt")) + 
  ggtitle("Ship")

s4_subset_float <- s4_subset %>% filter(type == "Float")

s4.ts3 <- PlotSvalbard::ts_plot(s4_subset_float, temp_col = "potT", sal_col = "s", color = "plot_date", symbol_size = 3, symbol_shape = 16, symbol_alpha = 0.5, color_isopyc = "black", WM = NULL) +
  scale_color_viridis_d(begin = 0.1, end = 0.8) +
  theme_classic2(18) +
  scale_x_continuous(name = "Practical Salinity", sec.axis = sec_axis(~.,name = expression(paste("Potential Density, kg m"^-3)))) +
  scale_y_continuous(name = "Potential Temperature, ˚C") +
   theme(panel.spacing.x = unit(1, "cm"),
        axis.text.x = element_text(angle = 0),
        legend.title = element_blank(),
        legend.key.size = unit(0.4, "cm"),
        legend.position = c(0.8, 0.25),
        legend.text = element_text(size = 12),
        legend.background = element_rect(size = 0.2, linetype = "solid", color = "black"),
        legend.spacing.y = unit(0, "pt")) +
  ggtitle("Float n0647")

s4.ts + s4.ts2 + s4.ts3

s6_subset <- floatsNmldsNezds %>% filter(Cruise == "AT38" & Station == 6, p <= 250,  datetime < ymd_hms("2017-09-18 00:00:00")) %>%  mutate(type = "Float") %>% select(-c(o2:poc)) %>% distinct() %>% 
  bind_rows(., ship_ctd %>% filter(Cruise == "AT38", p <= 250) %>% mutate(type = "Ship"))

s6_subset$plot_date <- reorder(s6_subset$plot_date, s6_subset$datetime)

s6.ts <- PlotSvalbard::ts_plot(s6_subset, temp_col = "potT", sal_col = "s", color = "type", symbol_size = 3, symbol_shape = 16, symbol_alpha = 0.7, color_isopyc = "black", WM = NULL) +
  scale_color_viridis_d(begin = 0.1, end = 0.8) +
  theme_classic2(18) +
  scale_x_continuous(name = "Practical Salinity", sec.axis = sec_axis(~.,name = expression(paste("Potential Density, kg m"^-3)))) +
  scale_y_continuous(name = "Potential Temperature, ˚C") +
   theme(panel.spacing.x = unit(1, "cm"),
        axis.text.x = element_text(angle = 0),
        legend.title = element_blank(),
        legend.key.size = unit(0.4, "cm"),
        legend.position = c(0.8, 0.7),
        legend.text = element_text(size = 12),
        legend.background = element_rect(size = 0.2, linetype = "solid", color = "black"),
        legend.spacing.y = unit(0, "pt"))

## Scale for 'x' is already present. Adding another scale for 'x', which will
## replace the existing scale.

## Scale for 'y' is already present. Adding another scale for 'y', which will
## replace the existing scale.

s6_subset_ship <- s6_subset %>% filter(type == "Ship")

s6.ts2 <- PlotSvalbard::ts_plot(s6_subset_ship, temp_col = "potT", sal_col = "s", color = "plot_date", symbol_size = 3, symbol_shape = 16, symbol_alpha = 0.7, color_isopyc = "black", WM = NULL) +
  scale_color_viridis_d(begin = 0.1, end = 0.8) +
  theme_classic2(18) +
  scale_x_continuous(name = "Practical Salinity", sec.axis = sec_axis(~.,name = expression(paste("Potential Density, kg m"^-3)))) +
  scale_y_continuous(name = "Potential Temperature, ˚C") +
   theme(panel.spacing.x = unit(1, "cm"),
        axis.text.x = element_text(angle = 0),
        legend.title = element_blank(),
        legend.key.size = unit(0.4, "cm"),
        legend.position = c(0.8, 0.7),
        legend.text = element_text(size = 12),
        legend.background = element_rect(size = 0.2, linetype = "solid", color = "black"),
        legend.spacing.y = unit(0, "pt")) +
  ggtitle("Ship")

## Scale for 'x' is already present. Adding another scale for 'x', which will
## replace the existing scale.
## Scale for 'y' is already present. Adding another scale for 'y', which will
## replace the existing scale.

s6_subset_float <- s6_subset %>% filter(type == "Float")

s6.ts3 <- PlotSvalbard::ts_plot(s6_subset_float, temp_col = "potT", sal_col = "s", color = "plot_date", symbol_size = 3, symbol_shape = 16, symbol_alpha = 0.7, color_isopyc = "black", WM = NULL) +
  scale_color_viridis_d(begin = 0.1, end = 0.8) +
  theme_classic2(18) +
  scale_x_continuous(name = "Practical Salinity", sec.axis = sec_axis(~.,name = expression(paste("Potential Density, kg m"^-3)))) +
  scale_y_continuous(name = "Potential Temperature, ˚C") +
   theme(panel.spacing.x = unit(1, "cm"),
        axis.text.x = element_text(angle = 0),
        legend.title = element_blank(),
        legend.key.size = unit(0.4, "cm"),
        legend.position = c(0.8, 0.7),
        legend.text = element_text(size = 12),
        legend.background = element_rect(size = 0.2, linetype = "solid", color = "black"),
        legend.spacing.y = unit(0, "pt")) + 
  ggtitle("Floats n0846 & n0847")

## Scale for 'x' is already present. Adding another scale for 'x', which will
## replace the existing scale.
## Scale for 'y' is already present. Adding another scale for 'y', which will
## replace the existing scale.

s6.ts + s6.ts2 + s6.ts3

Map

# chlaInfo <- info("erdMH1chla1day")
# n2_chl <- griddap(chlaInfo, latitude = c(46.5, 48.), longitude = c(-40.5, -37), time = c('2016-05-23','2016-05-27'), fields = 'chlorophyll')
# n3_chl <- griddap(chlaInfo, latitude = c(53.2, 53.5), longitude = c(-40, -39.4), time = c('2017-09-13', '2017-09-17'), fields = 'chlorophyll')


chlaInfo2 <- info("nesdisVHNSQchlaDaily")
n2_chl <- griddap(chlaInfo2, latitude = c(46.5, 48.), longitude = c(-40, -37), time = c('2016-05-23','2016-05-27'), fields = 'chlor_a')

## info() output passed to x; setting base url to: https://upwell.pfeg.noaa.gov/erddap/

n3_chl <- griddap(chlaInfo2, latitude = c(53.2, 53.5), longitude = c(-40, -39.4), time = c('2017-09-13', '2017-09-17'), fields = 'chlor_a')

## info() output passed to x; setting base url to: https://upwell.pfeg.noaa.gov/erddap/

sstInfo <- info("jplMURSST41")

n2_sst <- griddap(sstInfo, latitude = c(46.5, 48.), longitude = c(-40.5, -37), time = c('2016-05-24','2016-05-24'), fields = 'analysed_sst')

## info() output passed to x; setting base url to: https://upwell.pfeg.noaa.gov/erddap/

n3_sst <- griddap(sstInfo, latitude = c(53.2, 53.5), longitude = c(-40, -39.4), time = c('2017-09-13', '2017-09-13'), fields = 'analysed_sst')

## info() output passed to x; setting base url to: https://upwell.pfeg.noaa.gov/erddap/

library(ggnewscale)

custom.shapes <- c("NAAMES 2 Station 4" = 21, "NAAMES 3 Station 3" = 22, "NAAMES 3 Station 6" = 23)
custom.colors <- c("Ship" = "#377EB8", "Float n0847" = "#FF7F00", "Float n0846" = "#E41A1C", "Float n0647" = "#4DAF4A")
custom.lines <- c("Ship" = 1, "Float n0847" = 2, "Float n0846" = 2, "Float n0647" = 2)
mycolor <- colors$viridis
sst_color <- colors$temperature
w <- map_data("worldHires", ylim = c(35., 60.), xlim = c(-60, -30))


stations <- read_rds("Input/ctd_data.rds") %>% 
  select(CampCN:lat) %>% 
  distinct() %>% 
  group_by(CampCN) %>% 
  filter(row_number() == 1) %>% 
  rename(datetime = `yyyy-mm-ddThh:mm:ss.sss`,
         lon = `Longitude [degrees_east]`) %>% 
  ungroup()

map_data <- stations %>% select(Cruise, Station, Date, lon, lat) %>% mutate(type = "Ship") %>% 
  bind_rows(., floatsNmlds %>% select(Cruise, Station, float, datetime, lon, lat) %>% distinct() %>% rename(Date = datetime) %>%  mutate(type = paste("Float", float))) %>% 
  mutate(name = ifelse(Cruise == "AT34", "NAAMES 2", "NAAMES 3"), 
         plot_name = paste(name, "Station", Station),
         plot_sep = ifelse(Station == "6", "Subpolar", "Subtropical"))

map1_data <- map_data %>% 
  filter(Cruise == "AT34")

map1 <- ggplot(data = map1_data, aes(x = lon, y = lat)) + 
  facet_wrap(~plot_sep, dir = "v", scales = "free") +
  # geom_tile(data = n2_sst$data, aes(x = lon, y = lat, fill = analysed_sst)) +
  geom_tile(data = n2_chl$data, aes(x = lon, y = lat, fill = chlor_a), alpha = 0.7) +
  scale_fill_gradientn(colors = mycolor, na.value = NA) +
  labs(fill = expression(paste("Chlorophyll", italic(" a"), ", mg m"^-3))) +
  new_scale_fill() +  
  geom_line(aes(color = type, group = interaction(type, plot_name), linetype = type), size = 1) +
  geom_point(aes(fill = type, shape = plot_name),  size = 4, alpha = 0.9) +
  theme_bw() + 
  labs(x = "Longitude, ˚W", y = "Latitude, ˚N", color = "", fill = "", linetype = "", shape = "") +
  scale_linetype_manual(values = custom.lines) +
  scale_fill_manual(values = custom.colors) +
  scale_color_manual(values = custom.colors) + 
  scale_shape_manual(values = custom.shapes) +
  guides(fill = guide_legend(override.aes = list(shape = 21))) +
  ggtitle("NAAMES 2 Late Spring") 

map2_data <- map_data %>% 
  filter(Cruise == "AT38")

map2 <- ggplot(data = map2_data, aes(x = lon, y = lat)) + 
  facet_wrap(~plot_sep, dir = "v", scales = "free") +
  # geom_tile(data = n3_sst$data, aes(x = lon, y = lat, fill = analysed_sst)) +
  geom_tile(data = n3_chl$data, aes(x = lon, y = lat, fill = chlor_a), alpha = 0.7) +
  scale_fill_gradientn(colors = mycolor, na.value = NA) +
  labs(fill = expression(paste("Chlorophyll", italic(" a"), ", mg m"^-3))) +
  new_scale_fill() +  
  geom_line(aes(color = type, group = interaction(type, plot_name), linetype = type), size = 1) +
  geom_point(aes(fill = type, shape = plot_name),  size = 4, alpha = 0.9) +
  theme_bw() + 
  labs(x = "Longitude, ˚W", y = "Latitude, ˚N", color = "", fill = "", linetype = "", shape = "") +
  scale_linetype_manual(values = custom.lines) +
  scale_fill_manual(values = custom.colors) +
  scale_color_manual(values = custom.colors) + 
  scale_shape_manual(values = custom.shapes) +
  guides(fill = guide_legend(override.aes = list(shape = 21))) +
  ggtitle("NAAMES 3 Early Autumn")

map1 + map2 + plot_layout(guides = "collect")

POC profiles

There are only o2 profiles for May 24 and 25th for N2S4, so not enough data to track water mass using O2 or calc AOU via float

#subset the temperature and the mld data separately

poc <- floatsNmldsNezds %>% 
   filter(Cruise == "AT34", p <= 250,  datetime < ymd_hms("2016-05-31 00:00:00")) %>%
  select(profile, datetime, p, poc) %>% 
  mutate(decimaldate = decimal_date(datetime)) %>% 
  select(decimaldate, p, poc)

poc_points <- floatsNmldsNezds %>% 
   filter(Cruise == "AT34", p <= 250,  datetime < ymd_hms("2016-05-31 00:00:00")) %>%
  select(profile, datetime, MLD, p, poc) 

mba <- MBA::mba.surf(poc, no.X = 300, no.Y = 300, extend = T)
dimnames(mba$xyz.est$z) <- list(mba$xyz.est$x, mba$xyz.est$y)
mba <- reshape2::melt(mba$xyz.est$z, varnames = c('decimal_date', 'p'), value.name = 'poc')

Plot

FDOM profiles

#subset the temperature and the mld data separately

fdom <- floatsNmldsNezds %>% 
   filter(Cruise == "AT34", p <= 250,  datetime < ymd_hms("2016-05-31 00:00:00")) %>%
  select(profile, datetime, p, fdom) %>% 
  mutate(decimaldate = decimal_date(datetime)) %>% 
  select(decimaldate, p, fdom)

mba <- MBA::mba.surf(fdom, no.X = 300, no.Y = 300, extend = T)
dimnames(mba$xyz.est$z) <- list(mba$xyz.est$x, mba$xyz.est$y)
mba <- reshape2::melt(mba$xyz.est$z, varnames = c('decimal_date', 'p'), value.name = 'fdom')

Plot

Chl profiles

#subset the temperature and the mld data separately

chl <- floatsNmldsNezds %>% 
   filter(Cruise == "AT34", p <= 250,  datetime < ymd_hms("2016-05-31 00:00:00")) %>%
  select(profile, datetime, p, chla) %>% 
  mutate(decimaldate = decimal_date(datetime)) %>% 
  select(decimaldate, p, chla)

mba <- MBA::mba.surf(chl, no.X = 300, no.Y = 300, extend = T)
dimnames(mba$xyz.est$z) <- list(mba$xyz.est$x, mba$xyz.est$y)
mba <- reshape2::melt(mba$xyz.est$z, varnames = c('decimal_date', 'p'), value.name = 'chl')

Plot

chl.plot / poc.plot / fdom.plot

## Warning: Raster pixels are placed at uneven horizontal intervals and will be
## shifted. Consider using geom_tile() instead.

## Warning: Raster pixels are placed at uneven horizontal intervals and will be
## shifted. Consider using geom_tile() instead.

## Warning: Raster pixels are placed at uneven horizontal intervals and will be
## shifted. Consider using geom_tile() instead.